Trigonometry A

Average Accuracy: 99.8% → 99.8%

Time: 9.4s

Precision: binary64

Cost: 13376

\[0 \leq e \land e \leq 1\]

Math

\[\frac{e \cdot \sin v}{1 + e \cdot \cos v} \]

↓

\[\frac{e}{1 + e \cdot \cos v} \cdot \sin v \]

FPCore

(FPCore (e v) :precision binary64 (/ (* e (sin v)) (+ 1.0 (* e (cos v)))))

↓

(FPCore (e v) :precision binary64 (* (/ e (+ 1.0 (* e (cos v)))) (sin v)))

C

double code(double e, double v) {
	return (e * sin(v)) / (1.0 + (e * cos(v)));
}

↓

double code(double e, double v) {
	return (e / (1.0 + (e * cos(v)))) * sin(v);
}

Fortran

real(8) function code(e, v)
    real(8), intent (in) :: e
    real(8), intent (in) :: v
    code = (e * sin(v)) / (1.0d0 + (e * cos(v)))
end function

↓

real(8) function code(e, v)
    real(8), intent (in) :: e
    real(8), intent (in) :: v
    code = (e / (1.0d0 + (e * cos(v)))) * sin(v)
end function

Java

public static double code(double e, double v) {
	return (e * Math.sin(v)) / (1.0 + (e * Math.cos(v)));
}

↓

public static double code(double e, double v) {
	return (e / (1.0 + (e * Math.cos(v)))) * Math.sin(v);
}

Python

def code(e, v):
	return (e * math.sin(v)) / (1.0 + (e * math.cos(v)))

↓

def code(e, v):
	return (e / (1.0 + (e * math.cos(v)))) * math.sin(v)

Julia

function code(e, v)
	return Float64(Float64(e * sin(v)) / Float64(1.0 + Float64(e * cos(v))))
end

↓

function code(e, v)
	return Float64(Float64(e / Float64(1.0 + Float64(e * cos(v)))) * sin(v))
end

MATLAB

function tmp = code(e, v)
	tmp = (e * sin(v)) / (1.0 + (e * cos(v)));
end

↓

function tmp = code(e, v)
	tmp = (e / (1.0 + (e * cos(v)))) * sin(v);
end

Wolfram

code[e_, v_] := N[(N[(e * N[Sin[v], $MachinePrecision]), $MachinePrecision] / N[(1.0 + N[(e * N[Cos[v], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

↓

code[e_, v_] := N[(N[(e / N[(1.0 + N[(e * N[Cos[v], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] * N[Sin[v], $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 99.8%

   \[\frac{e \cdot \sin v}{1 + e \cdot \cos v} \]

2. Simplified 99.8%

   \[\leadsto \color{blue}{\frac{e}{\mathsf{fma}\left(e, \cos v, 1\right)} \cdot \sin v} \]
   Step-by-step derivation

   [Start] 99.8	\[ \frac{e \cdot \sin v}{1 + e \cdot \cos v} \]
   associate-*l/ [<=] 99.8	\[ \color{blue}{\frac{e}{1 + e \cdot \cos v} \cdot \sin v} \]
   +-commutative [=>] 99.8	\[ \frac{e}{\color{blue}{e \cdot \cos v + 1}} \cdot \sin v \]
   fma-def [=>] 99.8	\[ \frac{e}{\color{blue}{\mathsf{fma}\left(e, \cos v, 1\right)}} \cdot \sin v \]

3. Taylor expanded in v around inf 99.8%

   \[\leadsto \color{blue}{\frac{e}{1 + \cos v \cdot e}} \cdot \sin v \]

4. Final simplification 99.8%

   \[\leadsto \frac{e}{1 + e \cdot \cos v} \cdot \sin v \]

Alternatives

Alternative 1
Accuracy	99.6%
Cost	13248

\[\frac{\sin v}{\cos v + \frac{1}{e}} \]

Alternative 2
Accuracy	98.9%
Cost	6848

\[\sin v \cdot \frac{e}{e + 1} \]

Alternative 3
Accuracy	97.8%
Cost	6592

\[e \cdot \sin v \]

Alternative 4
Accuracy	52.8%
Cost	1344

\[\frac{e}{v \cdot \left(e \cdot -0.5 + -0.16666666666666666 \cdot \left(-1 - e\right)\right) + \left(\frac{e}{v} + \frac{1}{v}\right)} \]

Alternative 5
Accuracy	51.7%
Cost	576

\[\frac{e}{\frac{1}{v} + v \cdot 0.16666666666666666} \]

Alternative 6
Accuracy	51.2%
Cost	448

\[v \cdot \left(e - e \cdot e\right) \]

Alternative 7
Accuracy	51.7%
Cost	448

\[v \cdot \frac{e}{e + 1} \]

Alternative 8
Accuracy	50.6%
Cost	192

\[e \cdot v \]

Alternative 9
Accuracy	4.5%
Cost	64

\[v \]

Reproduce

herbie shell --seed 2023157 
(FPCore (e v)
  :name "Trigonometry A"
  :precision binary64
  :pre (and (<= 0.0 e) (<= e 1.0))
  (/ (* e (sin v)) (+ 1.0 (* e (cos v)))))